1. Field of the Invention
The present invention relates to an image processing method for resizing (expanding/reducing) image data to a determined desired magnification, and more particularly to a method of resizing photographic image data. The invention relates also to an image processing apparatus for implementing this method.
2. Description of the Related Art
Image data generated by an image data acquiring device such as a digital camera, a scanner vary in size significantly. For reproducing such data in the form of a hard copy, e.g. for obtaining a print of such data, there are generally defined so-called print sizes. Therefore, it is necessary to effect resizing (expansion/reduction) on the image data in advance so as to suit a certain desired print size. In doing this, in case the image data comprise photographic image data, basically the resizing operation is done in such a manner as to maintain its aspect ratio.
In the art of image reduction, e.g. the Japanese Patent Application “Kokai” No. 5-48881 (paragraphs [0005] to [0010] and FIG. 1) proposes a solution for simplifying hardware construction required therefor. In this, rather than reducing the image data to a desired magnification in a single stage, the reduction process is divided into two stages. In the first stage, the image data is reduced to ½n (n: an integer including 0), then, in the second stage, this image data reduced to 1/2n is further reduced to the desired magnification α ranging between 1 and ½, thus obtaining ( 1/2n)× α reduced image data. That is, in this technique, image data reduced in advance to 1/2n is reduced subsequently to the desired magnification α. Therefore, although this technique employs the apparently limited magnification of 1 to ½, by setting (n) to 0, a reduction ratio of 1 to 0.5 can be obtained; by setting (n) to 1, a further reduction ration of 0.5 to 0.25 can be obtained; and by setting (n) to 2, a still further reduction ratio of 0.25 to 0.125 can be obtained, respectively. As a result, this technique allows a wide range of reduction operation. In this, (first) resizing means for resizing to the intermediate magnification of 1/2n can be easily realized simply by using a shift register. Also, as second resizing means for resizing the image data to a desired final magnification of 1 to ½, this second resizing means needs to provide only such limited reduction ratio and needs to process only three pixels as reference pixels. Then, e.g. a coefficient generating circuit for effecting coefficient calculation according to each magnification can be simple, so that the entire circuit construction can be simple as well. However, this conventional image reducing technique does not contemplate any means for preventing occurrence of striped pattern on the outputted image, commonly known as “moire”. Hence, in the case of certain reduction ratios, moire appears conspicuously, which can result in image quality deterioration which is critical in the case of e.g. a photographic print in particular.
As an image processing technique attempting to cope with this problem, according to the Japanese Patent Application “Kokai” No. 11-298724 (paragraphs [0004] to [0005] and FIG. 1), first, a magnification: r designated by a magnification designating means is converted into separate elements: p−n and α, where p is a natural number, n is an integer including 0, and the coefficient α is 1≦α<p. Then, an image reducing means effects a reduction operation on image data inputted thereto for the reduction to p−n. Next, an image expanding means expands the reduced data to the magnification of α. This technique is based on a concept that occurrence of moire phenomenon can be avoided if the image reducing means effects an arithmetic averaging operation involving all pixels constituting the image data. However, even with this image resizing technique, it has been found that in the case of e.g. quality photographic print output, the moire phenomenon occurs, in certain magnifications, in such a degree as to result in image quality deterioration.